A Survey of the Literature July 29, 2010 Key Questions
1. Why should we care?
2. What do we know? 3. What can we know? Why should we care? • On one hand: o Households may need multiple sources of credit o Households may be perfectly able to manage their finances o On the other hand: o Multiple loans may reflect “cyclical debt” o Households may be unable to service their debt o Multiple borrowing may lead to rising levels of delinquency and default The words we use… Multiple borrowing = “Over-indebtedness” “Over-borrowing” “Multiple indebtedness” “Loan recycling” “Client poaching” “Loan pushing” “Credit pollution” “Over-supply of loans" The words we use…
…reflect our judgments about
• the cause(s) of multiple borrowing • (MFI side) • client poaching • loan pushing • (client side) • loan recycling • the consequences of multiple borrowing • over-indebtedness What do we know? • Multiple borrowing in other countries o Bangladesh o Bolivia o Bosnia and Herzegovina o India o Morocco o Nicaragua o Pakistan What do we know? • The few studies that exist employ various… 1) …data collection strategies o Consolidated MFI data o Aggregate data o Individual-level data o Qualitative approaches (e.g., “oral testimony”) 2) …and analytical approaches o Case studies o Descriptive statistics o Econometric techniques What do we know? • The incidence of multiple borrowing: examples Country Year Incidence Bolivia 2000 24 Bosnia and Herzegovina 2009 40 India (unnamed state) 2006 7 Morocco 2009 29 Nicaragua 2009 40 Pakistan 2009 21 Source: Chen and others (2010); Vogelgesang (2003); Krishnaswamy (2007) Incidence: Percent of borrowers with loans from more than one MFI What do we know? • The drivers of multiple borrowing o Collective behavior and tolerance for multiple borrowing o The role of economic shocks (health shocks) o “Risk-coping” versus “risk-prevention” o Lower-interest loans to repay other loans o Investment needs and existing average loan size What do we know? • The consequences of multiple borrowing: repayment behavior, default • Mixed evidence: examples o India: multiple borrowers perform as well as single borrowers o Bangladesh and Bolivia: multiple borrowers have a higher likelihood of irregular payment, default • Characteristics differ and may help explain outcomes • Context matters: worsening in crisis years What can we (not) know? • Some conceptual difficulties o Credibility of household-level information: How reliable are self-reported levels of indebtedness? o The unit of measurement: How do we treat individuals in the same household? o The sources of multiple loans: Should we include both formal and informal sources of finance? What can we (not) know? • The same observed outcome may reflect different underlying developments Household is • Unable to meet in financial repayment distress obligations Default Household is • Chooses not to repay selected obligations, with no not in financial impact on future ability to distress borrow What can we (not) know? • The same observed outcome may reflect different underlying developments Rising • Meets obligations but financial with rising difficulty distress At risk of Rising financial • Meets obligations default through “loan recycling” distress
Occasional • Seasonal difficulties in
financial meeting obligations distress What can we know? • Some possible strategies • Pooling MFI information and comparing with households’ self-reported level of indebtedness • (Static) Profile of borrowers with multiple loans • Consumption and investment behavior • Labor market activity • Repayment behavior • (Dynamic) Changes over time • The drivers of multiple borrowing • The impact on repayment behavior